DTV broadcasting in the United States of America has been done in accordance with broadcasting standards formulated by an industry consortium called the Advanced Television Systems Committee (ATSC), which standards have prescribed the use of a vestigial-sideband amplitude-modulated single carrier in each radio-frequency (RF) channel allocated for broadcasting DTV signals. Consideration is being given to replacing those DTV broadcasting standards with new standards that may prescribe coded orthogonal frequency-division multiplexed (COFDM) plural carriers in each RF channel allocated for broadcasting DTV signals. These new standards may, for example, resemble the DVB-T2 broadcasting standard developed for use in Europe.
COFDM is typically generated beginning with randomizing digital data to insure that subsequent encoding of forward-error-correction (FEC) coding receives sufficient density of logic ONEs to operate efficiently. Then, the resulting FEC coding is subjected to some form of bit interleaving, and the bits of the interleaved FEC coding are mapped to quadrature-amplitude-modulation (QAM) symbol constellations. The real-axis and imaginary-axis spatial coordinates of the QAM symbol constellations are parsed into orthogonal frequency-division multiplex (OFDM) symbols, which modulate a single carrier wave at high rate using quadrature-amplitude-modulation (QAM). The resulting modulated carrier wave is then transformed in a fast inverse discrete Fourier transform (I-DFT) procedure to generate a multiplicity of RF carrier waves uniformly distributed within the frequency spectrum of the RF channel, each of which RF carriers is modulated at low symbol rate. (In this specification and the accompanying drawing and claims the general term “QAM” is to be considered to include QPSK, QPSK being an alternative term for 4QAM.)
Reception of COFDM generated as described in the foregoing paragraph will fail if there is severe flat-spectrum fading of substantial duration. Such flat-spectrum fading is sometimes referred to as a “drop-out” in received signal strength. Such drop-out occurs when the receiving site is changed such that a sole effective signal transmission path is blocked by an intervening hill or structure, for example. Because the signaling rate in the individual OFDM carriers is very low, COFDM receivers are capable of maintaining reception despite drop-outs that are only a fraction of a second in duration. However, drop-outs that last as long as a few seconds disrupt television reception perceptibly. Automatic gain control of the front-end tuner stages of a DTV receiver will increase their gain, amplifying noise to introduce burst noise into the FEC coding. Such protracted drop-outs are encountered in a vehicular receiver when the vehicle passes through a tunnel, for example. By way of further example of a protracted drop-out in reception, a stationary receiver may briefly discontinue COFDM reception when receiver synchronization is momentarily lost during dynamic multipath reception conditions, such as caused by an aircraft flying over the reception site. Electric motors can generate radio-frequency noise strong enough to overload the front-end tuner stages of a DTV receiver, acting as a jamming signal that obliterates COFDM reception and generates burst noise too long to be corrected by FEC coding.
The DVB-T2 standard for DTV broadcasting prescribes Bose-Chaudhuri-Hocquenghem (BCH) coding concatenated with subsequent low-density parity-check coding (LPDC) as FEC coding. The concatenated BCH-LDPC coding prescribed in the DVB-T2 standard is reported to allow better performance in the presence of AWGN to be achieved using 256QAM symbol constellations than could be achieved with DVB-T using 16QAM symbol constellations. The bits of the LDPC coding are block interleaved using a modification of matrix type of interleaving in which successive bits of LDPC coding are arranged in columns for subsequent row-by-row utilization for mapping to lattice points within successive QAM symbol constellations, which medication introduces “column twist”. The DVB-T2 standard authorizes an alternative to parsing the real-axis and imaginary-axis spatial coordinates of 16QAM or 64QAM symbol constellations directly into orthogonal frequency-division multiplex (OFDM) symbols. In this alternative these QAM symbol constellations are in effect rotated relative to the real and imaginary axes of coordinate space so that every one of the lattice points of successive QAM symbols has unique coordinates along both the real axis and the imaginary axis. The unique coordinates of each QAM symbol constellation along the imaginary axis are then delayed for transmission respective to the unique coordinates of that QAM symbol constellation along the real axis. Accordingly, two unique coordinates of each QAM symbol constellation are transmitted, the second transmission being delayed respective to the first transmission for a time longer than the duration of a QAM symbol of an individual carrier. This facilitates iterative-diversity reception that can often avoid complete loss of reception of a portion of DTV signal owing to flat-spectrum fading, despite signal drop-outs that last up to a second or so.
An alternative way to facilitate iterative-diversity reception that can avoid complete loss of reception of a DTV signal owing to flat-spectrum fading, despite signal drop-outs that last up to a few seconds, is single-time retransmission of complete QAM symbol constellations after a suitable interval. Obviously, ignoring effects of signal corruption arising from noise in the transmission channel, twice transmitting complete QAM symbol constellations halves digital payload compared to transmitting them once.
When superficially considered, rotation of the QAM symbol constellations respective to the real and imaginary axes of modulation signal space appears not to reduce digital payload compared to a single-time transmission. Also, the respective times at which the initial transmission of a mapping of a set of data bits occur and at which the subsequent retransmission of a mapping of that same set of data bits occur are automatically referenced respective to each other.
However, the sizes of the data-slicing bins of the I-axis coordinates are reduced from what they would be for the square QAM symbol constellation without rotation. Also, the sizes of the data-slicing bins of the Q-axis coordinates are reduced from what they would be for the square QAM symbol constellation without rotation. E.g., the sizes of the data-slicing bins would be reduced by a factor of four when 16QAM symbol constellations are rotated. Reduction in the size of the data-slicing bins results in more bit errors being caused by AWGN corrupting the QAM. It is here pointed out that the reduced-size data bins are substantially the same size as those for non-rotated 256QAM symbol constellations. The number of bit errors in data slicing that are caused by AWGN corrupting the non-rotated 256QAM symbol constellations is substantially the same as the number of bit errors in data slicing that are caused by AWGN corrupting the rotated 16QAM symbol constellations. The number of data bits that can be mapped by each of the rotated 16QAM symbol constellations is four. The number of data bits that can be mapped by each of the non-rotated 256QAM symbol constellations is eight, and single-time retransmission of the non-rotated 256QAM symbol constellations results in the same eight data bits being conveyed over two COFDM symbol block intervals. Whichever of the two methods is used to transmit data twice, two COFDM symbol block intervals convey on average eight data bits times the number of QAM symbol constellations in each COFDM symbol block interval. I.e., for given size of data-slicing bins and given number of bit errors in data slicing that are caused by similar AWGN, single-time retransmission of 256QAM symbol constellations results in similar code rate as rotated 16QAM symbol constellations do. It appears that the developers of the DVB-T2 standard were unaware of this surprising result arising from more careful analysis.
When the Q-axis coordinates of rotated 16QAM symbol constellations are delayed respective to the their I-axis coordinates, the OFDM carriers are no longer each modulated in accordance with a respective rotated 16QAM constellation. Instead, each OFDM carrier is modulated in accordance with a respective 256QAM constellation. Presuming the 16QAM symbol constellations each used Gray mapping, the 256QAM symbol constellations are not Gray-mapped. The coordinates for the two orthogonal axes in which data-slicing is done are not Gray-coded, nor are they independent of each other. So, de-mapping rotated 16QAM constellations in a DTV receiver involves two-dimensional metrics for estimating errors in each of the four de-mapped bits to support subsequent soft decoding procedures for the FEC coding. This is a much more challenging task than de-mapping 256QAM symbol constellations that are Gray-mapped and have independent coordinates for the two orthogonal axes in which data-slicing is done. These independent coordinates are Gray-coded and de-mapping involves two sets of one-dimensional metrics for estimating errors in each of the eight de-mapped bits to support subsequent soft decoding procedures for the FEC coding. Errors can be estimated quite simply, proceeding from the departures of data-slicing results from values associated with lattice points in a 256QAM symbol constellation uncorrupted by noise.
Initial and final transmissions of the same coded data often allows a DTV receiver to replace burst noise in one of those transmissions with coded data from the other one of those transmissions. So, single-time retransmission of the same coded data facilitates the receiver being provided with substantial capability for overcoming peaks in noise of additive white Gaussian noise (AWGN) character, such as Johnson noise, as well as exceptionally effective capability for overcoming burst noise of substantial duration. This capability for overcoming burst noise facilitates the use of concatenated BCH-LDPC coding being used as FEC coding. LDPC coding is effective for overcoming AWGN, doing so at code rates almost three times those of concatenated convolutional coding (CCC) similarly effective for overcoming AWGN. LDPC coding does not have the tendency to extend burst errors that CCC has. However, LDPC coding is not very effective for correcting burst noise. While BCH coding can correct burst errors, using the BCH coding to correct a good amount of burst error distributed through each lengthy LDPC codeword requires many parity bits being associated with the systematic bits of the LDPC codeword. This undesirably reduces overall code rate of the concatenated BCH-LDPC coding. The DVB-T2 standard prescribes BCH coding capable of correcting only 10 or 12 bits in a block of 7,200 to 541,000 bits, keeping the number of parity bits per block less than 200. Accordingly, the retransmission provided by dissecting rotated symbol constellations is the principal mechanism allowing a receiver to correct burst noise in its reception of transmissions from a single COFDM transmitter.
If a reception site is not more than a few kilometers distant from the COFDM transmitter or transmitters, multipath reception can cause severe frequency-selective fading of a large group of OFDM carriers in a particular portion of the RF channel. Frequency-selective fading has been observed that extends over three MHz in a 6-MHz-wide RF channel and reduces the amplitude of a central few of the selectively faded OFDM carriers as much as 35 dB respective to less affected OFDM carriers. Such frequency-selective fading corrupts so many of the QAM symbols used for modulating respective OFDM carriers that de-interleaving of the results of de-mapping the QAM symbols is unable to reduce the density of bit errors in the recovered bit-wise FEC coding to permit successful decoding thereof. In some circumstances a directional reception antenna may be able to mitigate this problem. However, the antenna of a hand-held receiver is not apt to have appreciable capability for rejecting co-channel interference.